Stroke and dementia are not only common disorders of the elderly, but are also commonly associated. Cerebrovascular disease is considered the second most frequently cause of dementing illnesses. However, the magnitude of the risk of dementia posed by stroke, the mechanisms involved in the association, and the specific neuropsychologic features of the clinical syndrome of dementia in stroke compared to other neurobehavioral disorders are issues for further study. The first objective of this prospective two-phase study is to determine the incidence of dementia in patients with ischemic stroke. Phase I will document the prevalence of dementia in about 280 patients over age 60 with ischemic stroke admitted to the Columbia-Presbyterian Medical Center. Information on the neurologic and neuropsychologic deficits, functional ability, CT scan findings, stroke mechanism, and risk factors for vascular and Alzheimer's disease will be collected. From this group, a cohort of about 250 stroke patients free of dementia will be identified and followed annually over the four-year period of Phase II with detailed medical, psychiatric and neuropsychological assessments. A cohort of 250 comparison subjects comprised of healthy, non- demented, age-matched persons from the patient's neighborhood will undergo a similar study protocol. The relative risk of stroke- related dementia (SRD) will be calculated using survival analysis or life-table methods. The second objective, aimed at elucidating the mechanisms involved in SRD is to assess factors which may be associated with or predict the occurrence of dementia. These factors include CT abnormalities (e.g. atrophy, infarct number and volume); stroke mechanism (e.g. atherosclerotic, lacunar or cardioembolic), and risk factors related to vascular disease (e.g. hypertension) or Alzheimer disease (e.g. family history). Using univariate and multivariate techniques, a combination of factors will be evaluated, controlling for age, sex, and race. The third objective is to characterize the syndrome of dementia and other neurobehavioral syndromes in stroke subjects, aimed at defining a specific profile of defects which may clinically distinguish SRD from other forms of dementing illnesses including degenerative dementia. The longterm goal of this project is to improve the accuracy of the clinical diagnosis of stroke-related dementia and to provide the basis for preventive therapies.